|
Avid Pro Audio CommunityHow to Join & Post • Community Terms of Use • Help Us Help YouKnowledge Base Search • Community Search • Learn & Support |
#1
|
|||
|
|||
Latency?
I often use PT HD systems to record, but I am looking at the possibility of recording with native systems. I am trying to understand the latency.
I just tested the latency on my 002 on various buffer and low latency settings, I don't understand why there is such a big difference between the playback engine H/W buffer size and the actual latency (Changing the DAE buffer size appeared to make no difference). I measured the latency in two ways - with a mikeand percussive sound with the monitors feeding the signal back to give the most realistic scenario latency (and subtracting the time the sound takes to travel from the speakers to the mic) and also within PT - a source click sound chained via analogue outs and ins to a record channel and then on to another record channel. The results were consistent: H/W buffer measured latency 32 samples 269 samples 64 samples 402 samples 128 samples 605 samples 256 samples 1117 samples Low Latency 87 samples Low latency seems pretty good, but why is there such a big difference between the H/W buffer size and the actual latency? Will this latency be a lot smaller with other (non-HD) interfaces? Last edited by DaveUK; 02-19-2011 at 03:48 AM. Reason: I'd mistyped one of the figures in the table |
#2
|
|||
|
|||
Re: Latency?
Huh. I may need to test mine again. I tested my 002R in 2006 running whatever was current then - probably some flavor of 7.x. Came up with significantly less latency except with LLM, which is identical. See the results here.
What sample rate did you test at? That makes a big difference. Latency does vary some with different interfaces. It's all in the design of the converters and the hardware buffer. The buffer takes some of the load off the CPU somehow. Not sure why, but it does. I believe H/W buffers are built into the interfaces themselves. The DAE buffer is probably a software thing related to how the engine talks to the CPU or something. Totally different animal. Dang. Looks like I documented everything about my experiments except the version of PT I was running. I see from the mixing latencies page that I had EQ3, so that had to be at least 7.0, right?
__________________
David J. Finnamore PT 2023.12 Ultimate | Clarett+ 8Pre | macOS 13.6.3 on a MacBook Pro M1 Max PT 2023.12 | Saffire Pro 40 | Win10 latest, HP Z440 64GB |
#3
|
|||
|
|||
Re: Latency?
@daeron80,
Heh, I had come accross your results before but never realized you were the author! I think I found the page via a Google search, and did not read the footer with your info. Great write up.
__________________
| Logic Pro 9 | Reason 6.5 | Pro Tools 10 | MOTU UltraLite mk3 | RME ADI-2 | Summit Audio 2BA-221 x 2 | MOTU 8pre x 2 | | Mac Pro (4,1) 2.26gHz Xeon x 2 16gB OS X 10.7.4 | Macbook Pro (8,2) 2.2gHz i7 Quad 16gB OS X 10.7.4 | |
#4
|
|||
|
|||
Re: Latency?
Thanks!
__________________
David J. Finnamore PT 2023.12 Ultimate | Clarett+ 8Pre | macOS 13.6.3 on a MacBook Pro M1 Max PT 2023.12 | Saffire Pro 40 | Win10 latest, HP Z440 64GB |
#5
|
|||
|
|||
Re: Latency?
Hi Daeron, I hadn't seen your site before, very clear and great information.
I ran the tests on OSX, with the session at 48KHz. The figures seem to be in the same ballpark as yours, but offset by one - ie my 32 sample measurement corresponds roughly to your 64 sample measurement. My experiment (version one - not the mic test) seems to be similar to yours, except I measured the time by selecting a range visually zoomed in matching the first peak of my sample waveform (a cowbell sample) and took the readings from the sample ruler. |
#6
|
||||
|
||||
Re: Latency?
Don't forget; the buffer setting is only part of the total latency. There is also the conversion latency(somewhere around 1.2ms for a round trip).
__________________
HP Z4 workstation, Mbox Studio https://www.facebook.com/search/top/...0sound%20works The better I drink, the more I mix BTW, my name is Dave, but most people call me.........................Dave |
#7
|
|||
|
|||
Re: Latency?
I used your site (sent a link) to a guy I was in a band with, explaining that tracking latency was not really an issue. The problem with Pro Tools (at the time) was mixing delay compensation.
He was a really smart guy but the concept totally went over his head.
__________________
| Logic Pro 9 | Reason 6.5 | Pro Tools 10 | MOTU UltraLite mk3 | RME ADI-2 | Summit Audio 2BA-221 x 2 | MOTU 8pre x 2 | | Mac Pro (4,1) 2.26gHz Xeon x 2 16gB OS X 10.7.4 | Macbook Pro (8,2) 2.2gHz i7 Quad 16gB OS X 10.7.4 | |
#8
|
|||
|
|||
Re: Latency?
Quote:
In other words if the results were buffer size plus conversion delay + general system delay, we would expect the difference in latency to be only the number of extra samples added to the buffer. ie with a buffer of 128 we have a latency of 605 samples; one might imagine this was buffer+system+conversion delay, ie 128+477. Increasing the buffer to 256 we would then expect 256+477=733, but actually I measured 1117, a lot more. Its all academic really, the latency options are what they are, although I find the labels misleading. Do all DAWs have a similar sized discrepancy between buffer size and actual latency? |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Automation latency when using plugins that have long latency | pyrodave | Pro Tools 11 | 2 | 01-27-2014 11:22 AM |
omni/native latency vs mbox pro latency | chrisdee | Pro Tools HDX & HD Native Systems (Win) | 34 | 03-30-2012 07:24 AM |
Latency Issues - changing latency has no effect | Kippa-Dee | Pro Tools M-Powered (Win) | 1 | 07-15-2010 08:39 PM |
Piano VI's- Latency Latency!! | MARVINBASS | Virtual Instruments | 5 | 04-27-2006 01:09 AM |
Unity DS-1 and Latency... Anyone else feel the latency makes the Plugin unusable? | Mt.Everest | Pro Tools TDM Systems (Mac) | 7 | 08-26-2001 04:53 PM |